Conlang Random Thread

[bradrn]

Are there attested language with both implosives and ejectives at the same places of articulation, with the main contract being the mode of glottalic articulation? That is, as opposed to Hausa, which has implosive b and d but ejective ts and k.

Nevermind, there's Me'enit.

https://www.academia.edu/23703870/Sketc ... of_Me_enit

So ANADEW is a fair claim, I'd say.

Also Dahalo, Zulu, Yucatec Maya… It isn’t common, but it’s certainly well-attested.

[Kuchigakatai]

The fact that they dropped syllables gradually and irregularly from the table rather than, for example, just ditching whole rows all at once, suggests that it was never a perfect snapshot of the phonology of the language at the time of writing. Spelling lags pronunciation in many languages, perhaps a majority. If we assume that what was once /ə wə/ became /o wo/ and then both merged as /o~wo/ (an allophonic contrast, not phonemic), it seems reasonable to me to assume that the syllable we spell yo1 may have become identical with yo2 in pronunciation before it was dropped from the syllabary.

The gaps are perfectly phonetically reasonable though. Notice how the vowel pairs involving [j] (i1-i2, e1-e2) are missing the contrast after coronals and /w/ (the latter perhaps representing a merger of */wj/ with /jw/?), and the vowel pairs involving [w] (o1-o2) are missing after labials. Dropping [j] after a coronal (or palatal) and [w] after a labial are perfectly common sound changes.

This kind of thing is also attested elsewhere too. Most of North American English has dropped the /j/ of older /ju:/ after /t d/ and less often also /n/, producing the likes of "tune" /tʉn/ and "new" /nʉ/ (as opposed to SSBE /tju:n/ and /nju:/), but it is retained after most other consonants, as in "fume" /fjʉm/ or "cute" /kjʉt/. This is similar to the lack of [j] after /t d s z n r/ for Old Japanese /i e/.

Also, PIE as reconstructed didn't have much of /w/ as a glide inside an onset except after coronals (mostly s d dh, but also some t r) and laryngeals, plus a few cases after the so-called palatals (which as we know were probably velars). If you consider PIE kʷ gʷ gʷʰ as really "kw gw gʰw", the complete lack of /w/ after PIE /p bʰ/ is very similar to the lack of [w] after /p b/ for Old Japanese /o/.



Looking at what happened after onto Latin is relevant too (and also a bit fun). Proto-Italic remained much the same as PIE with regard to /w/ (minus the presence of laryngeals, which left bare instances of /w/), and further started treating the PIE labiovelars kʷ gʷ gʷʰ as the sequences /kw gw ɣw/. However, from Proto-Italic to Classical Latin, /w/ was eliminated after both coronals and dʰ ǵʰ > /f x/, when either type of consonant was followed by a short vowel:
- *dwis > bis
- *dwenos > bonus
- Old Latin dwellom > bellum
- *sweso:r > soror
- *swekrus > socrus
- *swenos > sonus
- *swepnos > somnus
- PIE *dʰworom > PI *fworom > forum
- PIE *dʰwo:r > (with unstressed shortening before sonorant) PI *fwor-a:-s > forās
- PIE *ǵʰweh₁ros > (with unstressed shortening before sonorant) PI *xweros > ferus

Exceptions: PI *quattwōr is a bit interesting because Latin just syllabified the -w- after -tt-: quattuor [ˈkwat.tu.ɔr]. (This also a common strategy to get rid of Cj: PI *spekjo: > speciō [ˈspɛ.ki.o:], that is, when it doesn't just drop the -j-, as in hesternus 'yesterday', or only leave the -j-, as in *pedjo:sem > peiōrem [ˈpɛjjo:rɛm].) Proto-Italic had already lost the [w] in *sweḱs because of *septm, otherwise we'd get Latin *sox for 'six'.

Otherwise, the /w/ is maintained when /sw/ is followed by a long vowel, or when it's after a Proto-Italic velar plosive. The latter includes the fortition of /ɣw/ to /gw/ after /n/ (otherwise /ɣw/ just becomes /w/: *ɣwenjo > veniō, *leɣwis > levis). (Apparently there are no descendants of /dw/ followed by a long vowel. Proto-Italic *dweiros > dīrus is likely a borrowing from Umbrian, but it would've given us unattested Latin *duīrus /ˈdwi:rʊs/ [ˈdɥi:rʊs].)

- *swa:dwis > suāvis [ˈswa:.wɪs]
- *swa:dtos > *swa:ssos > suāsus [ˈswa:sʊs]
- *swe:sko: > suēscō [ˈswe:.sko:]
- *linkwo: > linquō [ˈlɪŋ.kwo:]
- *linkwa:m > linquam [ˈlɪŋ.kwam]
- *kwa:lis > quālis [ˈkwa:.lɪs]
- *dngwa: > Old Latin dingua > lingua [ˈlɪŋ.gwa]
- *ekwo:i > equō [ˈɛ.kwo:]
- *ekwos > Old Latin equos [ˈe.kwos]
- *kwid > quid [kɥɪd] (with /Cw/ [Cɥ] allophone due to following front vowel)
- *kwie:ts > quiēs [ˈkɥi.e:s]
- *kwenkwe > quinque [ˈkɥɪŋ.kɥɛ]
- *sninɣwet > ninguit [ˈnɪŋ.gɥɪt]
- *ongwen > unguen [ˈʊŋ.gɥɛn]

Exceptions: PI *kwelō > colō. Interestingly, collum is probably not an exception, coming from *kolsom (< western IE *kolsom < *kʷol-es-om with mysterious but old derounding) or *kollom (< kokʷlom < kʷekʷlo(s)-m with dissimilation, shared with Baltic meaning 'neck').

So, basically, /w/ was lost everywhere except after the velars /k g/, except for a few words where a long vowel made it easier to say [sw] phonetically. And this is how Latin ended up with /kw gw sw/ as the only acceptable C + glide clusters (sometimes called /kʷ gʷ sʷ/).

Yes, i accept that /ɥ/ exists, but for the purposes of creating conlangs, I am going with my theory rather than setting up a language that has a /ɥ/ phoneme that only occurs in one specific environment, and indeed, in only one syllable in the whole language.

What I tried to convey last time when this topic came up was that phonological /jw/ shows up as [ɥ] at the surface. You see that a lot in diachronic Chinese, where /j/ is the initial and /w/ is the glide of the CGVC syllable structure, but that doesn't need to mean /ɥ/ is a separate phoneme that only appears in a very limited context.

[bradrn]

I’m making a conlang right now, and I’ve given it a fairly complex (for me) stress system:

All syllable have CV(C) structure. … Moras are defined as follows: a CV syllable is 1 mora, a CVː syllable (with a long vowel) is 2 moras, a CVC syllable (with a coda) is also 2 moras, and a CVːC syllable (with both a long vowel and a coda) is 3 moras. Words must be at least bimoraic. … Stress is predictable: the syllable with the most number of moras (i.e. the heaviest syllable) is stressed, so e.g. /ˈɬii.sə/, /ʔəˈtɬaaɣ/. If there are multiple heaviest syllables, the stress goes on the first, e.g. /ˈba.na/, /daˈlas.naː/. However, a syllable with /ə/ as its vowel may not be stressed; instead, the next heaviest (or next equal heaviest if there are multiple syllables with the same weight) syllable is stressed, so we get e.g. /ˈʔi.nəɣ/ rather than /ʔiˈnəɣ/, /da.ləsˈnaː/ rather than /daˈləs.naː/. Stress is not phonemic, but it may appear to be depending on syllabification, especially with prenasalised consonants: e.g. /wa.ⁿdal/ is stressed on the second syllable, whereas /wan.dal/ is stressed on the first syllable.

(Copied from the Romanization Challenge thread, where I’ve just written a phonological description of this language.)

Is this system of stress placement realistic?

[mèþru]

new unnamed conlang
the land where it is spoken in is called Dákwīn
all I know thus far is phonological and orthographical
I want to come up with a proto-language first that could evolve into the target phonology before I actually make any more developments
(p) t k (m) n (f) s ʃ h ɾ l (j) w
i u e o a
sounds in parantheses are only in borrowed words/roots
tones: macron is 455, acute is 24, caron 314, grave 412
syllable structure is CV(n) or kwV(n)
orthography is mix of logography and syllabrary á la cuneiform

[mèþru]

and I forgot all the voiced consonants in the inventory facepalms
(p) (b) t d k g (m) n (f) s z ʃ ɣ h ɾ l (j) w
gwV(n) is also allowed

[mèþru]

the proto-lang will already start with four tones, albiet less complex ones
I'm thinking something along the following:
t d k ɡ kʷ ɡʷ ʔ
n
s̪ z̪ s̠ z̠ x ɣ
r l w
ɪ ʊ e̞ ə ɔ ɐ ɑ

Any suggestions on how to get from A to B or how can 4 non-contour tones evolve into the system I described?

[akam chinjir]

If you start with just a low and a high, you can get falling by having the high tone spread to the right and merge with a low tone in the next syllable; and you can get rising tone by dropping coda ʔ after a low tone. You can also get both by losing peripheral vowels while preserving their tone. I guess your dipping tone would come from a HLH sequence.

(I'm assuming it can be left as a phonetic detail that you've got 455 rather than just 4 and 412 instead of just 42---unless you intend for them to occur only on trimoraic syllables?)

[Nortaneous]

so you have /t d k g kʷ gʷ ʔ n s̪ z̪ s̠ z̠ x ɣ r l w ɑ ɐ e ə o i u/ and you want to get to /t d k g kʷ gʷ n s z ʃ ɣ h r l w a e o i u/? there's hardly any distance in the consonants, so you could just merge the sibilants and have x > h. but maybe that's too easy

what's the time depth? probably pretty short, so at that point making a protolang might not be worth it. just have morphophonemes or w/e, the main thing you'll be getting is some vowel alternations from the presumable mergers

tones can generally do whatever but my impression is that asian-style contour tone systems tend to correlate with monosyllabism. could be a coincidence, but this also holds for iau; most of lakes-plain just has high low rising falling, iau probably developed more tones because it preserved disyllable tone contours but monosyllabized

[mèþru]

I was thinking that words will be usually disyllabic, maybe tri or mono. But that's actual words, not roots, and in any case I'm not committed to anything morphologically. I am however committed to having a writing system that's mainly a syllabrary with some logographs.

what's the time depth? probably pretty short

I was hoping for about 1500-2000 years. A lot of people are telling me I should turn codas into new tones, so I suppose that'd be a lot of the sound changes.

(I'm assuming it can be left as a phonetic detail that you've got 455 rather than just 4 and 412 instead of just 42

Yes, except if I were to write 455 as a single number I'd go with the 5, that's why it is doubled.

[mèþru]

1500-2000 years

Now that I think about it, I want it much, much earlier than that. I suppose I should redesign the proto-language.

[missals]

A little bit ago I imagined a language that had a case/number/definiteness and maybe gender system, but which only had two forms for each noun and only distinguished a single, highly-prominent cell in the paradigm.

Like say a language had a nominal suffix -ə, and it always appeared on nouns except the nominative singular indefinite form. So, if kot is '(a) dog', then kotə is 'the dog' the 'dogs' 'dog (obl.)' 'the dogs (obl.)' '(some) dogs' etc etc. I wonder if such a system exists in a language somewhere? it seems rather like the final stage in the erosion of an IE case/number/gender system.

Alternatively you have only the nominative definite singular marked, kotə 'the dog (nom.)', with kot being 'a dog' '(some) dogs' 'the dogs' 'dogs (obl.) etc

Hmm, or maybe it could be two different inflection classes. Like so...



I imagine all of this would be supplemented to some degree by inflection of articles and adjectives, like in German.

[Man in Space]

Yesterday (read: a few hours ago), I officially named the Tlar Kyanà ministry of intelligence: The Ê ê Ê Ê. Yes, that is the full name. Yes, they were all different etyma that collapsed to the same phonetic realization (had it existed in the protolanguage, it would have been originally *ʔjɛɾ B ɛʔ B ʔwɛ B ɛɾ B, all four words of which collapsed to → /e B2/).

[Pabappa]

The Ê ê Ê Ê.

*(breathes deep*)

Very nice choice, I love vowel-initial words.

[bradrn]

Yesterday (read: a few hours ago), I officially named the Tlar Kyanà ministry of intelligence: The Ê ê Ê Ê. Yes, that is the full name. Yes, they were all different etyma that collapsed to the same phonetic realization (had it existed in the protolanguage, it would have been originally *ʔjɛɾ B ɛʔ B ʔwɛ B ɛɾ B, all four words of which collapsed to → /e B2/).

I would be very interested in seeing a gloss of what each word in Ê ê Ê Ê actually means.

(I also can’t help but think that surely there’s a less ambiguous name for it…)

[Yalensky]

I wonder if such a system exists in a language somewhere? it seems rather like the final stage in the erosion of an IE case/number/gender system.

Not gender/case/number, but English verbs come to mind: I walk, we walk, you walk, they walk, he/she/it walks.

Yesterday (read: a few hours ago), I officially named the Tlar Kyanà ministry of intelligence: The Ê ê Ê Ê. Yes, that is the full name. Yes, they were all different etyma that collapsed to the same phonetic realization (had it existed in the protolanguage, it would have been originally *ʔjɛɾ B ɛʔ B ʔwɛ B ɛɾ B, all four words of which collapsed to → /e B2/).

I would be very interested in seeing a gloss of what each word in Ê ê Ê Ê actually means.

(I also can’t help but think that surely there’s a less ambiguous name for it…)

How about Tlar Kyanà calling it the "Four Ês"? (Man in Space, please don't say the word for "four" is also Ê.)

[alynnidalar]

A little bit ago I imagined a language that had a case/number/definiteness and maybe gender system, but which only had two forms for each noun and only distinguished a single, highly-prominent cell in the paradigm.

I quite like this, particularly the idea of having multiple classes where the "default" is different. So it's less about marking number/case/etc. and more about marking default vs. non-default.

[Ares Land]

A little bit ago I imagined a language that had a case/number/definiteness and maybe gender system, but which only had two forms for each noun and only distinguished a single, highly-prominent cell in the paradigm.

Like say a language had a nominal suffix -ə, and it always appeared on nouns except the nominative singular indefinite form. So, if kot is '(a) dog', then kotə is 'the dog' the 'dogs' 'dog (obl.)' 'the dogs (obl.)' '(some) dogs' etc etc. I wonder if such a system exists in a language somewhere? it seems rather like the final stage in the erosion of an IE case/number/gender system.

(...)

I imagine all of this would be supplemented to some degree by inflection of articles and adjectives, like in German.

That's a bit like Old French, and in fact a few nouns in Old French had paradigms much like what you suggest.
https://en.wikipedia.org/wiki/Old_French#Nouns

[Glass Half Baked]

This sort of thing (i.e. one distinct form and a bunch of identical ones) happens with verbs, too. In English, some verbs like "cut" only have two forms: third person singular present tense active voice indicative, and everything else.

[Pabappa]

I find it difficult to work with any language whose phonology is not to my liking. I created this wonderful language I called Oyster, and you can tell I put a lot of work into the diachronics because the phonology has lots of cute little gaps:

Code: Select all

                       PLAIN                      PALATALIZED

 Rounded bilabials:    pʷ  bʷ          w 
 Bilabials:            p   b   m                  pʲ  bʲ  mʲ 
 Alveolars:            t   d   n   s   l          tʲ  dʲ      sʲ   
 Postalveolars:        č   ǯ   ň   š   ł           
 Palatals:             ć   ǵ   ń   ś   y         (ć   ǵ   ń   ś   y)
 Velars:               k   ġ   ŋ   x   g          kʲ   
 Labiovelars:          kʷ  ġʷ      xʷ  gʷ         
 Uvulars:              q           h                         
 Rounded uvulars:      qʷ          hʷ

But I just dont like it. Its too messy. I want simple syllables. Fortunately I recently discovered that the Oyster language derives from a nation which would likely have spoken a different branch of the family than what I had built Oyster from, and it happens that even though the languages are closely related their phonologies are radically different, so my new sketch of Oyster has just

Code: Select all

 Labials:     p   b   m   f   v       
 Alveolars:   t   d   n   s       l
 Palataloids: č   ǯ   ň   š       y
 Velars:      k

Which I find much, much easier to work with. Even though in the new Oyster sketch the consonants I like are not that common (for example, /p/ rarely occurs in initial position), I am very happy that I was able to replace the beautiful but messy Oyster I sketch with the humble but easily pronounced Oyster II.
(The vowels are similar in both sketches.)

[Imralu]

English

Code: Select all

             singular     plural
absolutive   /dɒg/        /dɒgz/
possessive   /dɒgz/       /dɒgz/

Yes, we write them all differently, if we adhere to the prescriptive standard, but three of the four forms are pronounced identically with one, the default, being different.